Stippy is a CLI tool that can vectorize images using stippling, which can prove useful for technical artists using Plotters or Laser engraving. It is based on this paper.
In a virtual env run:
pip install -r requirements.txt
pip install -e .
In a shell, run:
python main.py input.jpg
You can control some parameters using the following flags:
-lr
controls the learning rate. Higher numbers means a potentially quicker convergence but migh also cause instability.-n
controls the number of iterations of the Lloyds algorithm to perform. Higher means yields a potentially better result at a cost of compute time.-p
controls the number of stippling points to use.-w
controls the number of workers to use in the KDTree.-inv
inverts the image.
For example:
python main.py input.jpg -lr 20 -n 100 -p 10000
Will compute 100 iterations of the Lloyds algorithm with 10000 stipple points and a learning rate of 20.